Estimating CO2 Emission Time Series with Support Vector Machines Regression, Artificial Neural Networks, and Classic Time Series Analysis

Abstract

Artificial intelligence machine learning has become very popular in recent years. It offers the ability to combine machine learning theory with many analyses such as classification, prediction models, natural language processing. Carbon dioxide emission is defined as the release of carbon, often caused by human nature, into the atmosphere. In the 19th century, the industrial revolution took place and the use of coal-powered industrial vehicles increased the amount of carbon released into the atmosphere. These gases released into the atmosphere have brought climate problems in proportion to the increase in temperature. Because of climate problems, the sweet water source of the earth’s ice pack continues to melt and the sea level rises. Therefore, the amount of carbon dioxide emission (metric tons per person) Artificial Neural Networks (ANN), Support Vector Machines Regression (SVMR), estimated by Box-Jenkins technique based on time series analysis and estimated estimates compared to MSE (mean square error) between 1990-2018. The comparison found that the Artificial Neural Networks have better predictive results on the SVMR and Box-Jenkins technique on the performance benchmark.